{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:VTHLECPVEU6LNGYOJAGUGXPT2E","short_pith_number":"pith:VTHLECPV","schema_version":"1.0","canonical_sha256":"acceb209f5253cb69b0e480d435df3d12662ba9e2e31e93be066174303ef366d","source":{"kind":"arxiv","id":"2402.16048","version":3},"attestation_state":"computed","paper":{"title":"How Likely Do LLMs with CoT Mimic Human Reasoning?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Cunxiang Wang, Guangsheng Bao, Hongbo Zhang, Linyi Yang, Yue Zhang","submitted_at":"2024-02-25T10:13:04Z","abstract_excerpt":"Chain-of-thought emerges as a promising technique for eliciting reasoning capabilities from Large Language Models (LLMs). However, it does not always improve task performance or accurately represent reasoning processes, leaving unresolved questions about its usage. In this paper, we diagnose the underlying mechanism by comparing the reasoning process of LLMs with humans, using causal analysis to understand the relationships between the problem instruction, reasoning, and the answer in LLMs. Our empirical study reveals that LLMs often deviate from the ideal causal chain, resulting in spurious c"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2402.16048","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-25T10:13:04Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"134ef387a3f7fae75500570b942e0b5472a6726caf97296937760e205920690e","abstract_canon_sha256":"60573fbb02a544e8c9c0366f5bd7cc26642646bb358b8377675f4945db2699af"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:48:12.636822Z","signature_b64":"otDy9GhKbwe1HujXv1ilY5qZ+jlMdd3aZlMOhbKUnUV748zYzzG132tYaP7JJWQPfpu1RBg6lQViQ8+pC4tzCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"acceb209f5253cb69b0e480d435df3d12662ba9e2e31e93be066174303ef366d","last_reissued_at":"2026-07-05T09:48:12.636232Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:48:12.636232Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How Likely Do LLMs with CoT Mimic Human Reasoning?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Cunxiang Wang, Guangsheng Bao, Hongbo Zhang, Linyi Yang, Yue Zhang","submitted_at":"2024-02-25T10:13:04Z","abstract_excerpt":"Chain-of-thought emerges as a promising technique for eliciting reasoning capabilities from Large Language Models (LLMs). However, it does not always improve task performance or accurately represent reasoning processes, leaving unresolved questions about its usage. In this paper, we diagnose the underlying mechanism by comparing the reasoning process of LLMs with humans, using causal analysis to understand the relationships between the problem instruction, reasoning, and the answer in LLMs. Our empirical study reveals that LLMs often deviate from the ideal causal chain, resulting in spurious c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.16048","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2402.16048/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2402.16048","created_at":"2026-07-05T09:48:12.636300+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.16048v3","created_at":"2026-07-05T09:48:12.636300+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.16048","created_at":"2026-07-05T09:48:12.636300+00:00"},{"alias_kind":"pith_short_12","alias_value":"VTHLECPVEU6L","created_at":"2026-07-05T09:48:12.636300+00:00"},{"alias_kind":"pith_short_16","alias_value":"VTHLECPVEU6LNGYO","created_at":"2026-07-05T09:48:12.636300+00:00"},{"alias_kind":"pith_short_8","alias_value":"VTHLECPV","created_at":"2026-07-05T09:48:12.636300+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2509.01535","citing_title":"CAT: Causal Attention Tuning For Injecting Fine-grained Causal Knowledge into Large Language Models","ref_index":2,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VTHLECPVEU6LNGYOJAGUGXPT2E","json":"https://pith.science/pith/VTHLECPVEU6LNGYOJAGUGXPT2E.json","graph_json":"https://pith.science/api/pith-number/VTHLECPVEU6LNGYOJAGUGXPT2E/graph.json","events_json":"https://pith.science/api/pith-number/VTHLECPVEU6LNGYOJAGUGXPT2E/events.json","paper":"https://pith.science/paper/VTHLECPV"},"agent_actions":{"view_html":"https://pith.science/pith/VTHLECPVEU6LNGYOJAGUGXPT2E","download_json":"https://pith.science/pith/VTHLECPVEU6LNGYOJAGUGXPT2E.json","view_paper":"https://pith.science/paper/VTHLECPV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.16048&json=true","fetch_graph":"https://pith.science/api/pith-number/VTHLECPVEU6LNGYOJAGUGXPT2E/graph.json","fetch_events":"https://pith.science/api/pith-number/VTHLECPVEU6LNGYOJAGUGXPT2E/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VTHLECPVEU6LNGYOJAGUGXPT2E/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VTHLECPVEU6LNGYOJAGUGXPT2E/action/storage_attestation","attest_author":"https://pith.science/pith/VTHLECPVEU6LNGYOJAGUGXPT2E/action/author_attestation","sign_citation":"https://pith.science/pith/VTHLECPVEU6LNGYOJAGUGXPT2E/action/citation_signature","submit_replication":"https://pith.science/pith/VTHLECPVEU6LNGYOJAGUGXPT2E/action/replication_record"}},"created_at":"2026-07-05T09:48:12.636300+00:00","updated_at":"2026-07-05T09:48:12.636300+00:00"}